Add link to paper

README.md

@@ -1,17 +1,17 @@
 ---
-license: cc-by-4.0
-pretty_name: Designed Vocalizations Dataset
 language:
 - en
+license: cc-by-4.0
+size_categories:
+- 100K<n<1M
 task_categories:
 - audio-to-audio
+pretty_name: Designed Vocalizations Dataset
 tags:
 - voice-conversion
 - sound-design
 - vocalization
 - speech
-size_categories:
-- 100K<n<1M
 configs:
 - config_name: raw
   data_files:
@@ -29,7 +29,7 @@ configs:
 
 # Designed Vocalizations Dataset
 
-**[Demo & audio samples](https://ncai-official.github.io/speech/publications/designed-vocalizations-dataset/index.html)**
+**[Paper](https://huggingface.co/papers/2607.20951)** · **[Demo & audio samples](https://ncai-official.github.io/speech/publications/designed-vocalizations-dataset/index.html)**
 
 The **Designed Vocalizations Dataset** supports voice conversion for *designed* vocalizations
 — monster growls, robotic voices, and other sound-designed timbres — an area left
@@ -210,4 +210,4 @@ If you use this dataset, please cite:
   booktitle = {Proc. Interspeech 2026},
   year      = {2026}
 }
-```
+```